Operational Solutions

• Built sensor integrations for MOD security contractor serving Heathrow Airport and decommissioned nuclear facilities
• Implemented MQTT-driven event processing with configurable thresholds for motion, light, and infrared sensors
• Designed for low-latency requirements - near-instantaneous event processing for security response
• Created standardised event republishing to trigger automated drone and ground vehicle deployment
• First production use of AI-assisted discovery - used ChatGPT to translate electrical engineering specs to development requirements
• Delivered complete sensor integration tested with physical hardware in home office environment

Operational Solutions is a Reading-based security company providing automated security for high-sensitivity sites - decommissioned Ministry of Defence facilities, former nuclear power plants, and commercial locations including Heathrow Airport. They engaged me for a three-month UK contract to integrate a new sensor vendor's hardware into their existing security ecosystem.

The technical challenge was integrating sensor hub units - devices that consolidate inputs from motion detectors, light sensors, infrared sensors, and other detection hardware - into an MQTT-driven event workflow. Each hub publishes sensor data via secured WiFi, and my implementation was a C# console application that processes these events, applies configurable thresholds, and republishes standardised events to trigger downstream responses. The downstream systems included automated drone deployment for investigation and automated ground vehicles for conditions where drone flight wasn't feasible.

Low latency was critical. When a sensor triggers, the response needs to be near-instantaneous - you can't investigate an intrusion event that occurred half a second ago. This constraint shaped every architectural decision, from message processing patterns to threshold evaluation logic.

The biggest challenge wasn't the code - it was the documentation. The sensor vendor provided engineering specifications written for electrical engineers, not software developers. This became my first production engagement where I used ChatGPT extensively for discovery, translating electrical engineering documentation into development requirements. While the actual integration code couldn't use AI assistance due to the restricted work environment, the AI-augmented discovery phase significantly accelerated my understanding of the hardware interfaces.

I had the sensor equipment on my desk at home for testing, which was genuinely fascinating - watching real hardware respond to the integration work in real-time. The project was delivered successfully and integrated into their production security systems.